Missing Routes

[Alan24178471]

The problem i have is when i install a route via a rpk file all the previously installed route vanish from the selection list but are still present in the route folder.

The only work around i have found is to uninstall a scenario to a route and reinstall it to bring the route back for selection.

Has anybody any ideas how i can get the routes back without all the unistalling and having to reinstall.

[colinlee]

Alan,
see my post "lost routes" dated 22 jan. It's on page 10 at the moment.
Regards
colin

[RSAdam]

Hi,
Although missing in the list in the front end, is the route/scenario actually being deleted from Rail Simulator? The answer is probably no.

We have seen this before and its a result of manually editing an xml file and accidentally leaving an error. This error can then get packaged up into an RPK that someone wants to share. If the error sits at the end of the XML file it is harmless, but when something is added underneath this error, that new data gets swallowed or encompassed by the error in the entry above it, thus making it invisible to Rail Simulator.

Trace back everything you have installed, adding each item one by one to find out where the error was first introduced.

[AndiS]

It would help all users if we had a clear list of the checks the RPK tool does, plus what the most frequent omissions by users are. I would have assumed that if I use the official tool to ship my creation, I could be sure that some sanity check is performed. What else would be the advantage over just zipping it and ship it with instructions how to hack the route.xml file?

When we praised the decision to use XML for KRS data, two years ago, we assumed that the times of unverified data would be gone. If this is not the case, maybe Mike can integrate such checks into TMB, just like he did in RouteRiter. Of course, first we would need documentation of the data format. If RSDL do not want to provide this, then they should provide solid checking software, IMHO. The first option comes at smaller labour cost for RSDL, but there seem to be some politics involved, so I am not suggesting anything. But leaving us with an official packaging tool which does not do all the sanity check that are technically possible at the time of shipping must lead to user frustration.

[CaptScarlet]

Adam, is this something that could be remedied by either a 3rd party utility or Railsim.exe ( for instance ) rebuilding that xml file from the installed content if necessary ? Perhaps another option would be for the package manager creating a backup of the file when installing content ?

John

[RSderek]

Hi,
Somewhere along the line Users have to take resposiblity for their own content and what they package up.
The package manager like everything else to do with RS is being worked on and improved.
DoesWin zip check for shonk?

AndiS I can send our bank details on if you want to take the hit for the small labour cost.


regards
Derek
p.s. I may have to lay off the coffee.

[RSAdam]

It is interesting to note the instant jump in the direction of RSDL being at fault...

[CaptScarlet]

It is interesting to note the instant jump in the direction of RSDL being at fault...

Not from me. I just try to find solutions

John

[CaptScarlet]

Adam and or Derek, you might find this article/post interesting http://www.quartertothree.com/game-talk ... hp?t=42663 .

John

[AndiS]

AndiS I can send our bank details on if you want to take the hit for the small labour cost.

Just send the docu instead, then people around the world will take a major strike at reducing your labour cost.

My point was: We know that ZIP does nothing to check the shipped content. We did not know that the Kuju packager does the same. For ZIP, it is clear, how could that know about KRS. For a tool that comes with KRS, some people might have got the impression that they might rely on it at least for the simpler checks, like those that Adam seems to refer to -- XML file which are not correct at the basic level.

John, for the routes.xml file the answer is yes and no. This file contains not only the reference to the folder, and the name of the route (in 5 languages), but also references to a route blueprint, several sky blueprints, weather blueprint, details on the map projection (origin of the route), etc. Therefore, the thing to do would be to keep regular backups and paste together a new routes.xml file once it gets destroyed. That can be done in any editor, like TMB. Just avoid XML-Notepad, as that puts something into the file that serz.exe does not digest.

What I did when I lost my (benign) test routes due to my carelessness (no backup = sloppy user) was to create a new, empty route and then copy everything from the old (now "invisible") route folder to the new one. This was quick an painless.

In that process, I discovered that the random filenames which KRS uses for the route folders are the same for every install. This can cripple your merge effects as on two installs of KRS, two different routes will have the same folder name. This is a well known problem in computer generated "random" numbers and the well known remedy is to use the current time as part of the random number generation "seed". However, it was not known in Surrey at that time, obviously.

P.S.: Adam, what jump, whose fault?
Unfortunately, we do not know details of what exactly went wrong in Alan's case, but the first jump I saw was someone underlining the statement that users often edit XML files and do that erroneously. Now one of the well known properties of the XML standard is that they are easily verified for syntactic correctness by standard tools, provided you have the DTD. In the case of KRS, this DTD is hidden from the public, but one would assume that it is encoded somewhere in the programme. So the user is not a bad guy pointing at innocents if he assumes that XML be checked by software. If it is not, I would point at Kuju first, not RSDL, btw., just if you are interested in that.

For a positive look into the future, just consider what I wrote above, and we can forget about laying the blame on anyone.

[RSderek]

We not blaming anyone, just too much coffee.
John, that was one of the best things I've read for a long time.
I had to ask around the office to make sure none of us wrote it...

[CaptScarlet]

I thought you might like it , it was written by THQ director of creative management Michael Fitch.

John

[Retro]

Hi,
I have not used the package manager yet for uploads but have installed Stock and Items from UK.TrainSim using it. I have not installed any third party routes because of the possiblities of an XML mess up. I am also getting to the stage in a month or so of uploading my own Route. Looking at the topics created by the users of the aforementioned routes. The developer is bombarded with the multitude of posts asking why things won't work, files are missing and Routes vanishing puts me off using the Package Manager. I accept that maybe some people are not using it correctly but surely the software should be able to check for known problems that occur such as faulty Route XML Files or if it has the correct files it needs so the Route can be installed easily at the other end by the user. On The Sims 2 a Clean Installer has been produced for checking as it installs or packages Lots (houses) for items that are known to cause huge problems to the Game Software or corrupt files.
It would be nice to be able to rely on the KRS Package Manager in the same way.
This is not intended to be critical of anyone are RSDL just my opinions as a user of the Program.
Regards James

[karma99]

In that process, I discovered that the random filenames which KRS uses for the route folders are the same for every install. This can cripple your merge effects as on two installs of KRS, two different routes will have the same folder name. This is a well known problem in computer generated "random" numbers and the well known remedy is to use the current time as part of the random number generation "seed". However, it was not known in Surrey at that time, obviously.

Just wanted to correct this, when you install RS you get your default routes which always have the same ID's.. 0000-0000-0000-0001,2,3,4 etc, if I remember correctly.
When you create a new route (from within the game) the new route will have an ID created for it. This ID is a GUID which is not random but which will ALWAYS be unique (it's based on all kinds of internal pc guff and timings etc) and has bazillions(tm) of combinations so we can never create new routes using RS which will have clashes with routes made by anyone else, or ourselves.
I thought this was a great idea and eliminates issues with 3 people creating LondonToBrighton routes and the overwriting chaos that can ensue

[AndiS]

Thanks for the pointer, John. It might well be the case that us 100 or 500 innocent train simulation freaks are just a dying species in a sea of short-lived game (ab)users. This would explain a large part of the communication difficulties between Kuju, RSDL and the forums.

At any rate, as a constructive suggestion for the next patch or upgrade or whatever, I propose that the packaging part of the packager be modified to check as much as your code time allows and that the installing part be modified to a) make a backup of all global files it modifies, e.g., routes.xml or scenarios.xml, and b) checks the format of the resulting global file and if it is faulty, or contains less routes or scenarios for any reason, then the original is moved by, and the flawed post-install version is stored under another name, and a full description of what went wrong is output. Then, and only then, will we be able to shed some light into the entangled responsibilities of uploader, downloader, and software manufacturer.

@karma99, this is my memories of what I did: I had one install with the default routes plus one made by myself. After I corrupted it in some way (by my own doings), I moved it to the side and restored a copy of the plain installation, with only the default routes. Then I created a new route there, and that had the same folder name as the first route in the old installation. At least this is what I remember of the incident. Next time I crash something, I will take care to record the details, but I hope it will not be too soon.